Optimization of rotary compressor

Abstract

The following report documents the selection of an meta optimized optimization algorithm to be used in the optimization of 8 design variables to achieve the best performance in compressor simulations. The Complex RF optimization algorithm was selected due to its low computing costs, high convergence rate and its capabilities as a meta heuristic method capable of optimizing itself. Prior to optimization of compressor simulations, 5 different equations were selected to test the capabilities and characteristics of the selected optimization algorithm before using these 5 equations together to perform meta optimization on the selected algorithm. Meta optimization of the Complex RF method resulted in the meta optimized performance tuners of 4.6 for reflection coefficient, 0.006 for randomization factor and 0.111 for forgetting factor. A link was also established between the optimization algorithm and mathematical models in compressor simulation to find for the best weighted sum of mechanical efficiency, volumetric efficiency and coefficient of performance. Using the meta optimized factors, the Complex RF method was performed on the compressor simulations using the weighted sum of compressor simulation efficiencies and coefficient of performance. Results from the parametric study of design variables were then used to verify the results obtained from the optimization of the compressor simulation and then further improve the performance of the compressor by identifying constraints which can be adjusted.